Movement and behaviour of large southern bluefin tuna ... · tional double-barbed nylon dart tip...

Movement and behaviour of large southern bluefin tuna(Thunnus maccoyii) in the Australian region determinedusing pop-up satellite archival tags

TOBY A. PATTERSON,* KAREN EVANS,THOR I. CARTER AND JOHN S. GUNN

CSIRO Marine and Atmospheric Research, GPO Box 1538,Hobart, Tas. 7001, Australia

ABSTRACT

Pop-up satellite archival tags (PSATs) were deployedon 52 large (156–200 cm length to caudal fork)southern bluefin tuna (Thunnus maccoyii) in the wes-tern Tasman Sea during the austral winters of 2001–2005. Southern bluefin tuna (SBT) were resident inthe Tasman Sea for up to 6 months with movementsaway from the tagging area occurring at highly vari-able rates. The data indicated a general tendency forSBT to move south from the tagging area in theWestern Tasman Sea. Four individuals migrated westalong the southern continental margin of Australiaand into the Indian Ocean. Three individuals movedeast into the central Tasman Sea, with one individualreaching New Zealand. We also describe the firstobserved migration of an SBT from the Tasman Sea tothe Indian Ocean spawning grounds south of Indo-nesia. Individuals spent most of their time relativelyclose to the Australian coast, with an estimated 84%of time spent in the Australian Fishing Zone. SBTfavored temperatures between 19 and 21�C, adjustingtheir depth to the vertical temperature distribution.Distinct diurnal diving patterns were observed andadjustment of depth to maintain constant ambientlight levels over a 24-h period. The findings of thisstudy are a significant advance toward greater under-standing of the spatial dynamics of large SBT andunderstanding the connectivity between distantregions of their distribution.

Key words: habitat preferences, pop-up satellite archivaltag, southern bluefin tuna, spatial dynamics, spawningmigration

INTRODUCTION

Southern bluefin tuna (Thunnus maccoyii) are a largelong-lived pelagic predator widely distributed throu-ghout the oceans of the Southern Hemisphere (Caton,1991; Clear et al., 2000) and the focus of a large, high-value, multi-national, commercial fishery throughouttheir range. Genetic evidence indicates a single dis-crete spawning region, and the distribution of larvaesuggest southern bluefin tuna (SBT) are a single pop-ulation with spawning limited to an area in the north-eastern Indian Ocean south of Indonesia (Caton, 1991;Grewe et al., 1997; Farley and Davis, 1998). Adult SBTare assumed to forage throughout the temperate watersof the Southern Hemisphere oceans during the australwinter, migrating to the spawning grounds of thenorth-western Indian Ocean from spring to autumn(Shingu, 1978; Caton, 1991) before returning to for-aging grounds in the following autumn⁄winter. Indi-viduals do not remain on the spawning grounds overthe whole season; instead there is a turnover of fishwith the numbers of mature fish peaking in Octoberand February (Farley and Davis, 1998).

Stock assessments suggest that the current popu-lation of SBT is at 5–12% pre-exploitation biomass(CCSBT, 2004). However, many key inputs to man-agement are still unknown. Amongst these is con-siderable uncertainty about the spatial dynamics andtiming of spawning migrations. Spawning groundcatch data indicate a higher proportion of youngerfish in recent years relative to the 1990s, suggestingpossible changes in the age distribution of spawningSBT (Farley and Davis, 2005). Stock assessmentmodels assume a knife-edge recruitment of SBT to thespawning stock at the age of 10+ and implicitlyassume that the spawning stock is composed of obli-gate spawners. Although recent studies suggest thatrecruitment to the spawning stock is more diffuse,occurring from age 10 or older (Davis and Farley,2001; Schaefer, 2001), few data are available on thefrequency of spawning in mature SBT, the fidelity offish to foraging regions or mixing rates of SBTbetween these grounds. Current assumptions aboutthe spatial dynamics of mature SBT are derived from

*Correspondence. e-mail: [email protected]

Received 23 October 2007

Revised version accepted 23 June 2008

FISHERIES OCEANOGRAPHY Fish. Oceanogr. 17:5, 352–367, 2008

352 doi:10.1111/j.1365-2419.2008.00483.x � 2008 The Authors.

interpretation of longline catch data and conven-tional mark-recapture data studies, mostly on juve-niles. Catch data are thought to be biased by spatialcontraction and temporal variability in effort andchanges to targeting practices (Toscas et al., 2001)and more recently, subject to potential bias due tounder-reported catches (CCSBT, 2006). Additionally,conventional tagging relies on accurate reporting ofrecapture information and minimal non-reportingof recaptures (Hearn et al., 1999). The developmentof pop-up satellite archival tags (PSATs), whichtransmit data from the fish without the need for thetag to be recovered, provide fishery-independentmethods for assessing movement in pelagic fish (Gunnand Block, 2001).

Although this technology has been widely utilizedon other bluefin species (genus Thunnus) (Lutcavageet al., 1999; Block et al., 2005; Wilson et al., 2005),the deployment of PSATs on SBT has been limited.Four tags deployed on large SBT in the region of thespawning grounds north of 20�S (Itoh et al., 2002)demonstrated movements south from the spawningground over periods of a few days to 3 months. Threeindividuals undertook south-westerly movements intothe Indian Ocean and the fourth moved into an areasouth of the Australian continent. Given the smallnumber of releases and the short tag-attachmentdurations of those PSATs deployed, it is difficult toinfer much regarding the movement patterns of largeSBT.

Critical gaps currently exist in our knowledge andunderstanding of the movements, residency, regionalfidelity and spawning dynamics of older SBT. Thesegaps inhibit accurate assessment of stock status andalso management within regional fisheries. Here wereport on the results of pop-up satellite archival tag-ging of large SBT in the Tasman Sea and their dis-persal from these waters into the Southern Ocean,providing an important first step toward greaterunderstanding of the spatial dynamics of SBT in theAustralian region.

METHODS

Tagging operationsPop-up satellite archival tags (PAT2: N = 11, PAT3:N = 8 and PAT4: N = 33; Wildlife Computers, Red-mond, WA, USA) were deployed on large SBT in thewaters of the western Tasman Sea in the austral win-ters of 2001-2005. Fish were caught during commerciallong line operations, with those in good condition ledinto a tagging cradle and then lifted on board thevessel. Females sampled from the spawning ground

demonstrate 50% maturity at a length of 154 cm1

(Farley and Davis, 1998); we therefore chose only fishgreater than 154 cm length to caudal fork (LCF) fortagging.

In 2001 and 2002 tags were rigged with a singletitanium anchor connected via a 400-lb monofilamenttether to the corrodible release pin of the PSAT. In2003–2005 nylon umbrella-style anchors (Domeieret al., 2005) were used as primary anchors. An addi-tional double-barbed nylon dart tip crimped to a 50-lbmonofilament loop was attached as a secondary anchorto further secure the PSAT and to minimize any lat-eral tag movement. The primary monofilament leaderon all tags was fitted with a depth-release device (RD-1500, RD-1800; Wildlife Computers) designed toprevent implosion of the tag at great depth. The pri-mary anchor was inserted into the dorsal musculatureat the base of the second dorsal fin following severalother studies (Lutcavage et al., 1999; Stokesbury et al.,2004). The secondary anchor was inserted into thedorsal musculature in line with the dorsal finlets. Eachtag was printed with reward and return information.The deployment position was recorded using thevessel’s onboard GPS system.

Data and analyses

We programmed PSATs to record pressure (depth),temperature and light at 60-s (N = 50) or 120-s(N = 2) intervals. To provide a mix of higher resolu-tion, short-term data specific to management appli-cations (see Hobday and Hartmann, 2006) and lowerresolution longer-term data capable of capturingbroader movements, tags were programmed to releasefrom the fish after 30 days (N = 3), 60 days (N = 4),90 days (N = 1), 180 days (N = 7), 300 days (N = 1)and 365 days (N = 35). After release, tags floated tothe ocean surface and transmitted a summary of theirarchived data via the Argos satellite service (ServiceArgos, Toulouse, France). Due to limited transmissionbandwidth, data collected by the PSATs were sum-marized into 1-h (N = 1), 4-h (N = 22), 8-h (N = 20)and 12-h (N = 7) time periods prior to transmission.The summary data for each time period consisted ofdistributions of the proportion of time-at-depth(TAD) and time-at-temperature (TAT) and temper-ature-depth profiles. For those PSATs recovered, thefull archived dataset was downloaded from the tag. Toavoid possible behavioural changes imposed by theprocess of tagging, only those data collected from tags

1Schaefer (2001) quotes the figure 164 cm FL. We conser-

vatively chose the lower of these.

Movement of large southern bluefin tuna 353

� 2008 The Authors, Fish. Oceanogr., 17:5, 352–367.

at liberty for more than 14 days were included inanalyses.

Age classification of SBT

To determine whether tagged SBT would be classifiedas recruited into the spawning stock (as per recentmanagement assumptions), we derived age estimatesfor each individual using the cohort slicing method.This method is used in Commission for the Con-servation of Southern Bluefin Tuna (CCSBT) stockassessments (Preece et al., 2004) and assigns an in-dividual to an age cohort using length-at-age growthcurves (Laslett et al., 2002).

PSAT attachment duration

The rate of premature tag shedding was examined bynormalizing all tag deployments to a nominal start dayand plotting the proportion of tags remaining attachedto SBT as a function of days attached. We fittedintercept-free regressions of the form:

lnðproportion tags shedÞ ¼ l� days attached

to examine the shedding rate.

Location estimates

Daily positions derived from each tag were calculatedusing a heuristic algorithm (cf., Teo et al., 2004;Shaffer, 2005) that combined light and temperaturedata. Longitude was estimated using proprietary soft-ware (WC-GPE.1.02.0000; Wildlife Computers).Daily estimates of sea surface temperature (SST) werederived from SSTs reported by the tag where the depthof the SST observation was less than 5 m. These werechecked for outliers (i.e., extreme temperatures> 35�C) and erroneous measurements were replacedby linear interpolation using the 3 points either side ofthe gap. A loess smoother (Venables and Ripley,1997) was applied to the SST to reduce variability inthe SST series. Remotely-sensed SST values usingMCSST⁄AVHRR data (PO.DAAC, 2001) for 0.5� oflongitude either side of each longitude estimate werethen compared to SST values reported by the PSAT.Sea surface temperature values between 50�S and 10�Sfor each longitude estimate were gridded into cells of0.25� latitude and the number of matches to theobserved PSAT SST was counted in each latitudinalcell. Matches were defined as instances when the tagSST was within 0.1�C of the remote sensing SST. Toreduce the number of unrealistic positions generatedby spatially-distant matches to SST values, we con-strained the distance an individual could move eachday by weighting the latitudinal cell frequencies by afunction assumed to represent the likelihood of a

movement between consecutive positions. A lognor-mal distribution

Prðdistance;l;rÞ ¼ ðxrffiffiffiffiffiffi

2ppÞ�1 expð� lnðx� lÞ2=2r2Þ

was used to model the probability of distance moved.This weighting allows longer distance movements butwith lower weight than shorter distance movements.The parameters l and q2 were set at the values l = 4and q2 = 1. Using the formula for the expectationfrom the lognormal distribution, E(distance) =exp(l + ½ q2), the highest weightings are given tomovements of approximately 150 km day)1, orapproximately 0.97 body lengths s)1 in accordancewith published swimming speed estimates for a 170-cmtuna (Bushnell and Jones, 1994; Lutcavage et al.,2000). The latitude cell with the combination of thehighest number of matches and weights was consid-ered to be the most probable position. This process wasinitialized at the known tag-release position. Weevaluated the results of our method by comparing thegeolocation estimates with transmitted pop-up loca-tions where the last estimated position was no morethan 2 days prior to the first transmission date of eachPSAT. To derive the minimum distances traveled, wecalculated the distance of a simplified and idealizedtrack from the tagging release area to the pop-up pointutilizing the ‘tracker’ tool in the GIS software MAN-IFOLD 6.50 (Manifold Net Ltd., Carson City, NV,USA). This approximates the minimum ‘biologicaldistance’ that must be traveled by the SBT from thetagging locations to the spawning grounds.

Habitat preferences and behaviour

Aggregate time-integrated indices of temperature anddepth preferences were constructed by calculating themedian TAD and TAT value in each histogram bin.Empirical cumulative distribution functions were cal-culated from the median proportion in each bin andused to estimate the expected proportion of time in agiven depth or temperature range. These were mat-ched to location estimates to examine spatial patternsin depth and temperature preferences.

RESULTS

A total of 52 pop-up satellite archival tags weredeployed on SBT 156–200 cm LCF (mean ± SD:173.7 ± 9.5 cm – all estimates hereafter are mean ±SD unless otherwise stated) estimated to be aged 9–20 yr(15.3 ± 3.1 yr; Table 1). The cohort slicing methodutilized in CCSBT stock assessments assigned 50 ofthe 52 individuals tagged to the spawning stock. Ofthe 52 PSATs deployed, data were retrieved from 44

354 T.A. Patterson et al.


Table 1. Release and pop-up details for PSATs deployed on SBT in the western Tasman Sea 2001-2005.

Tag

Releases Pop-up transmissions

DateLatitude(�S)

Longitude(�E)

LCF(cm)

ProgrammedDuration

Age(yr) Date

Latitude(�S)

Longitude(�E)

200128703 13 Jul 2001 33.42 151.55 170 300 14 16 Jul 2001 33.38 151.8528701 13 Jul 2001 35.88 151.57 175 30 16 19Jul 2001 35.90 153.0228707 13 Jul 2001 35.22 151.53 190 60 20 07 Aug 2001 37.32 154.1228708 13 Jul 2001 35.23 151.53 158 60 10 27 Jul 2001 37.43 156.6213274 15 Jul 2001 35.58 151.60 178 180 19 20 Jul 2001 35.17 152.4313275 13 Jul 2001 35.60 151.60 200 180 20 13 Aug 2001 33.72 153.12200228709 27 Jul 2002 35.02 151.65 157 30 10 20 Aug 2002 35.07 152.6813272 27 Jul 2002 35.08 151.65 165 60 12 06 Sep 2002 36.82 152.1020926 27 Jul 2002 35.12 151.67 173 90 15 Failed to transmit200313279 10 Jul 2003 35.98 150.96 171 30 14 06 Aug 2003 34.71 152.2713273 10 Jul 2003 35.98 150.94 187 60 20 07 Sep 2003 33.72 153.3220890 10 Jul 2003 35.97 150.85 200 180 20 03 Aug 2003 30.81 153.8030466 10 Jul 2003 35.97 150.85 186 180 20 14 Oct 2003 35.55 154.7520914 20 Jul 2003 36.82 150.77 175 180 16 Failed to transmit20924 28 Jul 2003 35.05 151.80 160 364 10 Failed to transmit30465 28 Jul 2003 35.05 151.80 168 364 13 24 Jan 2004 42.59 123.1818564 28 Jul 2003 35.10 150.82 178 180 19 1 Jan 2004 38.16 122.6720891 07 Aug 2003 34.23 151.77 165 364 12 16 Sep 2003 30.60 154.96200420925 27 Jun 2004 34.18 151.87 176 180 17 21 Oct 2004 40.73 149.1243935 05 Jul 2004 34.13 152.80 174 364 16 13 Dec 2004 42.80 153.3743936 05 Jul 2004 34.16 152.88 173 364 15 15 Jul 2004 30.41 155.7943937 05 Jul 2004 34.18 152.90 182 364 20 04 Sept 2004 33.45 157.8643925 12 Jul 2004 34.67 153.15 175 364 16 26 Jul 2004 32.55 154.6143926 12 Jul 2004 34.67 153.15 169 364 13 03 Feb 2005 17.72 111.0743927 12 Jul 2004 34.67 153.15 183 364 20 14 Jul 2004 33.97 155.2443928 13 Jul 2004 34.65 153.27 170 364 14 14 Jan 2005 34.87 111.9043929 13 Jul 2004 34.65 153.27 174 364 16 08 Nov 2004 43.25 150.3143931 13 Jul 2004 34.65 153.27 169 364 13 19 Nov 2004 42.55 163.8843932 13 Jul 2004 34.63 153.27 169 364 13 15 Jul 2004 33.59 155.0543933 14 Jul 2004 34.68 153.23 176 364 17 Failed to transmit – 172 days data

recovered43934 15 Jul 2004 34.62 153.15 173 364 15 29 Jul 2004 34.51 152.1143945 30 Jul 2004 34.97 151.98 171 364 14 13 Dec 2004 44.60 146.6043946 30 Jul 2004 34.92 151.98 189 364 20 Failed to transmit43943 31 Jul 2004 35.00 151.95 169 364 13 15 Nov 2004 44.61 158.6943941 08 Aug 2004 34.70 152.72 172 364 15 29 Dec 2004 41.46 144.1243942 08 Aug 2004 34.82 152.95 173 364 15 15 Jan 2005 44.24 146.2643939 29 Aug 2004 36.42 152.90 175 364 16 27 Nov 2004 40.80 133.5943940 28 Aug 2004 36.37 152.87 170 364 14 30 Dec 2004 42.95 149.35200553269 07 Jul 2005 34.48 151.55 170 364 14 27 Jul 2005 34.51 153.3953270 21 Jul 2005 32.24 153.82 177 364 18 Still at liberty43944 21 Jul 2005 32.89 153.68 173 364 15 10 Dec 2005 43.58 147.7846336 22 Jul 2005 32.40 153.59 163 364 11 29 Oct 2005 40.54 137.0846337 22 Jul 2005 32.59 153.55 171 364 14 13 Sep 2005 33.32 155.1146339 22 Jul 2005 32.63 153.53 172 364 15 21 Sep 2005 37.21 154.0246340 01 Aug 2005 35.13 152.15 187 364 20 02 Sep 2005 34.67 153.43



(84.6%; Table 1) via Service Argos with dataretrieved from one recovered PSAT which failed totransmit. A total of two PSATs were recovered afterwashing up on land.


All PSATs detached prematurely, transmitting 1–362 days prior to the programmed deployment day.Tag attachment durations varied considerably bothwithin and between deployment years, ranging from 2to 206 days, with considerably better tag retention inyears where secondary anchors were used (Fig. 1). Theproportion of the intended deployment periodachieved ranged from 0.01 to 0.98 (Table 2). Trendsin the overall rate of tag loss (Fig. 1) were described bya negative exponential relationship, ln(proportion tags

attached) = l ·(days attached) (adjusted R2: 0.97,F1,39 = 1,535, P < 0.001). This model estimated theslope coefficient to be )0.02 (SE: ± 0.001) and anattachment ‘half-life’ (the number of days at which50% of the tags are likely to remain attached) of35.4 days (Fig. 1).

Geolocation accuracy

Adequate light and SST data were available for thecalculation of position estimates for 40 SBT. Of these,36 were at liberty for periods greater than 14 days andposition estimates calculated constituted 5.2–92.6% ofthe total time tags were at liberty (50.7 ± 21.3%).Comparison of position estimates calculated viageolocation with a final pop-up position derived fromArgos was restricted to 25 PSATs, due to a lack of SST

Table 1. Continued.

Tag

Releases Pop-up transmissions

DateLatitude(�S)

Longitude(�E)

LCF(cm)

ProgrammedDuration

Age(yr) Date

Latitude(�S)

Longitude(�E)

46342 19 Aug 2005 35.68 153.60 168 364 13 20 Oct 2005 43.91 147.6446343 19 Aug 2005 35.68 153.60 185 364 20 21 Nov 2005 41.17 150.5346350 19 Aug 2005 35.68 153.58 178 364 19 27 Dec 2005 41.60 140.9246351 21 Sep 2005 36.22 152.52 157 364 9 23 Sep 2005 34.82 153.1146353 21 Sep 2005 36.13 152.40 156 364 9 Failed to transmit53263 21 Sep 2005 36.11 152.34 164 364 11 06 Feb 2006 38.33 140.6753264 22 Sep 2005 35.97 153.05 172 364 14 Failed to transmit

(a)

(b)

Figure 1. (a) Normalized tag loss ratesby deployment year. (b) The overallnormalized rate of tag loss through time.The dashed lines indicate the ‘half-life’of the PSAT (time till 50% of tagsdetatched). Tag data from 2002 was notincluded due to small sample sizes.



data for latitude calculation, compromised light dataresulting in highly erroneous longitude calculation, ordata transmission failure. The median distance orposition error (± SD) between the two final positionswas calculated to be 161.9 ± 230.6 km (range 23.1–866.3 km, N = 25).

Migration and residency patterns

Individuals were resident in the western Tasman Seafrom June through to December, predominately in anarea bounded by 30–40�S and 150–160�E (Figs 2 and3). The period spent within this region ranged from 34to 55 days, with individuals moving out of the regionas early as September and as late as December, al-though most individuals had moved out of the regionby October. Five individuals undertook movementseast into the central Tasman Sea, with one moving tosouth-western New Zealand coastal waters beforereturning to the western Tasman and finally movingeast of Tasmania. Two SBT returned to the westernTasman Sea before moving south to an area east ofTasmania (Figs 2 and 3), with one transiting to thenorth-west of Tasmania. PSATs on the remaining twodetached before movement out of the Tasman Seacould be discerned.

Considerable time was spent west and north-west ofTasmania (Figs 2 and 3). Movements into this regionfrom the deployment area occurred as early as Sep-tember and as late as January. The time taken to reachthe area north-west of Tasmania from the deploymentarea averaged 26.8 ± 19.5 days (range 3–66 days).

Individuals were widely dispersed throughout thesouthern margins of Australia during the summermonths. Fish that moved into the western Australianarea were aged 13–19 yr (14.8 ± 2.9 yr) and ranged insize from 168 to 178 cm LCF (173.7 ± 9.6 cm). Theseindividuals reached western Australian waters atvarying times throughout November and December,taking 36–84 days to reach the area after departing the

western Tasman Sea. Of those SBT tracked into thewaters south-west of Australia, one travelled into whatis regarded as the region of the spawning grounds at17.72�S, 111.07�E (Fig. 3), whereas two SBT madeeasterly movements (possibly indicating initiation ofeastward return movements) after periods of residencysouth-west of Australia (Figs 2 and 3). By estimatingthe total distance from the western Tasman Sea to thearea of the spawning grounds using an idealized trackaround the Australian continent and Tasmania, theSBT which migrated to the spawning ground traveleda distance of � 9000 km in 113 days, resulting in anaverage movement rate of � 80 km day)1. Movementof the SBT migrating along the western coast ofAustralia from an area south of Cape Leeuwin atapproximately 36�S to the final pop-up position on thesouthern spawning grounds totaled 1550 km traveledover 26 days. This is equivalent to an average distancetraveled of 59.6 km day)1.

Although limited by small sample sizes, no rela-tionship between the size of individual SBT and theextent of their westward movements was apparent. It isalso interesting to note that the majority of positionestimates (84%) were in the Australian Fishing Zone(AFZ) and thus relatively close to the coast and shelf.Also, there was little indication of residency in theGreat Australian Bight (GAB) – an area of largejuvenile SBT aggregations during the summer months(Caton, 1991).


Time spent at depth by tagged SBT largely reflectedtemperature preferences of < 18–20�C (Fig. 4), withthe majority of time spent at < 200 m and the highestproportion of time spent at < 50 m. The PDT datashowed that SBT experienced minimum and maxi-mum ambient water temperatures of 2.6–30.4�C(Fig. 4) with a median maximum temperature of 18�C(inter-quartile range: 15.4–19.4�C) and a median

Table 2. Deployment periods and attachment durations achieved by PSATs deployed on SBT in the western Tasman Sea2001–05.

Year (N)

Deployment period (days) Attachment period (days) Proportion of period attached

Mean ± SD Range Mean ± SD Range Mean ± SD Range

2001 (6) 135.0 ± 103.5 30–300 16.2 ± 46.1 2–31 0.22 ± 0.13 0.01–0.422002 (3) 60.0 ± 30.0 30–90 32.5 ± 66.1 24–41 0.74 ± 0.08 0.68–0.802003 (9) 211.3 ± 127.3 30–364 83.3 ± 63.3 24–180 0.57 ± 0.36 0.10–0.982004 (20) 354.8 ± 54.1 180–364 99.1 ± 12.0 3–206 0.30 ± 0.20 0.01–0.642005 (14) 364.0 ± 0.0 364 69.7 ± 11.5 2–142 0.21 ± 0.13 0.01–0.39

Total 30–364 70.8 ± 61.0 2–206 0.33 ± 0.25 0.01–0.98



minimum temperature of 15�C (inter-quartile range:12.2–17.4�C). The cumulative distribution functionof time-at-temperature from all individuals pooled,estimated that SBT occupy waters of < 18.5�C for30% of the time, at or below 20�C for 50% of the timeand at or below 21�C for 90% of the time observed(Fig. 4c).

Individuals spent time in waters deeper than 200 mand cooler than 20�C primarily in three localizedareas: the western Tasman Sea centered on thedeployment area, the eastern Tasman Sea and theregion between 135 and 145�E north-west of Tasma-nia (Figs 5 and 6). Time spent in waters > 20�C wasmostly near the eastern Australian coast in the Tas-man Sea and on the spawning grounds in the IndianOcean (although these data come from a singleindividual). The fine scale depth and temperature datafrom the two recovered PSATs reflect the depth and

temperatures experienced from transmitted PSATdata (Fig. 7). Depth data recovered from both tagsshowed that SBT spent greater than 90% of their timein waters shallower than 250 m. Relative to the otherindividuals, PSAT 43933 spent considerably moretime at depths deeper than 400 m and colder than10�C during the summer when it was resident off thewest coast of Tasmania.

Spawning ground habitat data

The SBT that migrated to the spawning ground(PSAT 43926) traversed several different water massesduring its migration. Whilst in the western TasmanSea, temperatures were largely restricted to 15–20�Cand waters less than 200 m depth (Fig. 7). On initi-ating southward movement, the SBT encounteredvariable surface temperatures ranging from 10 to20�C. Throughout the waters south of Australia, time

(a)

(b)

(c)

Figure 2. (a) Deployment locations ofPSAT-tagged SBT in the western Tas-man Sea. (b) Deployment-to-popupdisplacements (c) position estimates ofSBT at liberty between 2001 and 2006Coloured by month.



spent in waters deeper than 250 m increased and, inassociation, temperatures experienced decreased, withwaters of 10–15�C experienced more frequently(Fig. 7). The depth and temperature experienced bythis individual shifted just prior to reaching tropicalwater masses in mid-January, with increased time insurface waters (< 150 m) punctuated by dives greaterthan 500 m. Sampled temperature in the region of thespawning grounds was typified by warm waters up to28�C.

Diurnal depth preferences

Lengthy periods of diurnal diving behaviour weredisplayed in the recovered archival data (Fig. 8).Whereas depths frequented during the day were quitevariable (ranging from 150 to 600 m), both SBTfavored waters less than 50 m at night during theseperiods. Diurnal diving behaviour associated with timespent in waters deeper than 400 m, was observed forperiods of up to 2 weeks at a time and resulted in SBT

spending periods of over 10 h in water temperaturesless than 10�C and as low as 7�C. Periods of diurnaldiving behaviour resulted in SBT remaining at more orless constant ambient light levels (Fig. 8).

DISCUSSION

This study provides the first substantive, fishery-inde-pendent, observations of the movement of large SBT.The duration of residency of individual SBT, bothwithin and beyond the western Tasman Sea, washighly variable, as was the extent of movements. Thedata suggest that migration schedules can be highlyplastic in nature and could depend on a complex offactors including the status of energetic reserves. Therelationship between energetic reserves and conditionhas been examined by Golet et al. (2007) who foundsignificant declines in lipid content over a decadalperiod in Atlantic bluefin tuna, resulting in possibleasynchronous maturity schedules.

(a)

(b)

(c)

Figure 3. Individual tracks of SBT forwhich PSATs remained attached for> 14 days separated into regions ofmovement and residency outside of thewestern Tasman Sea region: (a) centraland western Tasman Sea; (b) SouthernOcean and (c) Southern⁄Indian Oceans.



Depths and temperatures experienced by SBT werevariable, although spatially coherent, suggesting thatthermal preferences may be driven by proximate factorssuch as forage availability rather than thermal toler-ances. This agrees with Schick et al. (2004) who foundthat the locations of schools of Atlantic bluefin were notrelated to the locations of fronts but instead may havebeen influenced by unobservable prey distributions.Care must be taken in drawing conclusions about thosefactors driving SBT movement and behaviour.

Although there was a general pattern of movementsouth from the tagging area, around Tasmania and intothe Southern Ocean, many tags detached prematurely,resulting in a reduced capability of describing move-ment and habitat beyond the Tasman Sea region.


The problem of achieving long-term PSAT attach-ments was the single largest hindrance to accumulat-ing long-term data on the movements and habitatpreferences of SBT. To attain a more comprehensiveunderstanding of movement patterns of large SBT,improved attachment durations are critical. Incorpo-ration of a secondary anchor into deployments coin-cided with a considerable increase in attachmentduration, a finding supported by a number of otherstudies utilizing similar secondary anchors (H. Dewar,National Marine Fisheries Service, La Jolla, CA,USA, personal communication). However, theincreased attachment durations coincident with the

(a)

(b)

(c)

Figure 4. The average proportion of time spent at (a) depthand (b) temperature, and (c) the cumulative distributionfunction of the proportion of time spent at temperature forsouthern bluefin tuna tagged in the western Tasman Sea2001–2005. The median proportion of time spent withineach histogram bin is given in blue.

(a)

(b)

(c)

15°C

15–20°C

20°C

Figure 5. The proportion of time spent in temperatures (a)< 15�C, (b) between 15 and 20�C and (c) in temperatures> 20�C in relation to the estimated position of SBT taggedin the western Tasman Sea 2001–2005.



use of our secondary anchors are correlative and notcausative. Other factors, such as the expertise of thetagger or variability in the reliability of the tags, mayhave also contributed to attachment durations.

Determining whether the performance of attach-ments achieved in this study are comparable to thoseachieved elsewhere on bluefin species is difficult due to

a lack of published data; commonly the achievedattachment duration is published but the intendeddeployment time is not (e.g., Stokesbury et al., 2004).However, failure of tags to remain attached untilprogrammed release dates has been widely reportedelsewhere (Domeier et al., 2003; Gunn et al., 2003;Horodysky and Graves, 2005; Wilson et al., 2005) andis a common problem in studies utilizing PSATs.Maximum attachment durations achieved in this studywere considerably shorter than those reported else-where [206 days in this study, compared to 371 days inStokesbury et al. (2004), 304 days in Wilson et al.(2005) and 261 days in Block et al. (2005)]. A numberof factors may contribute to reduced attachmentsincluding anchor and tether design, tag attachmentmethods and⁄or physiological and behavioural differ-ences between species resulting in differences in tissuerejection of anchors or greater wear on attachmentpoints and tethers.

Geolocation accuracy

The errors calculated around the final PSAT positionestimates in this study compare well to the estimates ofgeolocation error calculated elsewhere (e.g., Teo et al.,2004; Sibert et al., 2006a) and indicate that themagnitude of errors are sufficiently small for inferenceof large-scale movements. Comparison of the endpoints of the track with Service Argos pop-up posi-tions may well be optimistic. Tags floating at thesurface of the ocean are likely to be easier to geolocatecompared to a tag attached to a tuna. However, we areconfident that conclusions about the timing ofmigrations and duration of residency in various areasare valid at the large scale we consider, despite beingobviously noisy estimates of the true movement path.Further improvement to statistical geolocation meth-odology (e.g., Royer et al., 2005; Lam et al., 2008) islikely to increase the accuracy of position estimatesand allow for more detailed characterization ofmovement and habitat usage. Importantly, advanceson current geolocation methods would allow for sta-tistically rigorous characterization of the uncertaintyin the location estimates.


Individuals tagged in this study clearly demonstrated apreference for waters at 18–20�C, similar to thatobserved in juvenile SBT (Gunn and Block, 2001) andother bluefin species (Block et al., 2001; Marcineket al., 2001; Stokesbury et al., 2004; Kitagawa et al.,2006). However, large SBT can clearly withstand muchlower temperatures for sustained periods, spendingperiods of over 10 h at temperatures less than 10�C.

(a)

(b)

(c)

(d)

(e)

Figure 6. The proportion of time spent in depth ranges (a)0–50 m, (b) 50–200 m (c) 200–400 m (d) 400–600 m and(e) 600–1000 m in relation to the estimated position of SBTtagged in the western Tasman Sea 2001–2005.



Therefore thermal preferences are unlikely to solelyreflect the physiological tolerance of SBT, but mightreflect the physiological limitations of forage species.Tuna are widely documented to associate with frontsand transition zones (Laurs and Lynn, 1977; Olson,2001; Polovina et al., 2001; Royer et al., 2004). How-ever, the factors driving this association are still largelyunknown (Kirby et al., 2000). Frontal areas are oftenassociated with biomass maxima, concentratingchlorophyll production and associated secondary pro-ductivity. Gunn and Young (1999) hypothesized thatSBT use warm-core eddies and the warm side of fronts asthermal refuge after periods of foraging in colder waters.Use of such oceanographic features to trade-off forageavailability against thermoregulatory requirements hasbeen postulated for other species of tuna (Neill et al.,1976; Sund et al., 1981), whereas others have hypoth-esized that this association is strictly related to theaggregation of forage species only (Brill and Lutcavage,2001; Brill et al., 2002; Royer et al., 2004). Whether ornot this association has links to thermoregulatory

requirements, the long forays into cool water demon-strated by SBT in this study represent a physiologicalcapability which may allow this species to capitalize onprey concentrations (when they are present) by maxi-mizing the amount of time spent in these frontal andeddy regions. Diurnal patterns in diving behaviourobserved in tuna species, including bigeye tuna (Musylet al., 2003), Atlantic bluefin tuna (Lutcavage et al.,2000; Schick et al., 2004; Gutenkunst et al., 2007) andPacific bluefin tuna (Kitagawa et al., 2001) have simi-larly been associated with the movement of foragespecies either through direct tracking of prey or trackingof particular light levels which may be easiest forambushing or detecting prey (Warrant, 2000; Brillet al., 2002). Similar behaviour has been observed inseveral other species such as broadbill swordfish (Careyand Robison, 1981) and bigeye thresher sharks (Nakanoet al., 2003; Weng and Block, 2004) and has been pos-tulated to be related to vertical migration dynamics ofthe deep-scattering layer. The changes in divingbehaviour on a diurnal scale associated with relatively

(c) (d)

(b)

(a)

Figure 7. Hourly mean temperature in5-m depth bins from full archival recordsretrieved from PSATs deployed onsouthern bluefin tuna in the westernTasman Sea: (a) tag 30466 and (b) tag43933. (c) Temperature and depth dataand (d) monthly temperature depthprofiles of SBT 43926 which migratedfrom foraging grounds in the westernTasman Sea to the area of the spawninggrounds in the Indian Ocean.



constant light levels, as observed in this study, suggestthat light may play an important role in determining theforaging depth of SBT.

Behaviour on the spawning ground

Although limited, the data obtained from the oneindividual tracked to the spawning grounds providesan interesting insight into the behaviour of SBT bothduring migration to the spawning grounds and whilstin the area of the spawning grounds.

The diving behaviour of this individual differedmarkedly between regions, with the fish demonstratingrelatively consistent diving behaviour ranging down to500 m whilst in the Southern Ocean, much deeperdiving behaviour in the Indian Ocean and surface-related diving behaviour in the region of the spawninggrounds. This cessation of deeper diving and a tran-sition to surface-related behaviour appeared to beassociated with the occurrence of surface temperaturesgreater than 24�C. Such surface-related behaviour issimilar to that documented by other SBT in thespawning ground region (Itoh et al., 2002) and is

consistent with suggestions that spawning occurs insurface waters (Davis and Farley, 2001). Additionally,it appears consistent with the vertical movementsobserved in spawning Atlantic bluefin (Teo et al.,2007). Reasons for the change from occupying depthsshallower than 500 m to much deeper diving whenmoving from the Southern to the Indian Ocean areunclear but may be associated with changes in preyspecies targeted and in associated changes in preyspecies distributions and⁄or changes in the thermalproperties of the water masses experienced. The dee-per diving demonstrated by this SBT (and, in associ-ation, the increased time spent in cooler waters) mighthave been avoidance of warm tropical surface water.Teo et al. (2007) found that spawning Atlantic bluefinin the Gulf of Maine displayed deep diving en route tospawning regions preceding a period of shallow diving.This was hypothesized to be either a thermoregulatoryresponse or a means to reduce energetic demands byavoiding stronger currents. However, much more dataare required to characterize the vertical distribution ofSBT in this region.

(a)

(b)

Figure 8. (a) Instances of diurnal divingbehaviour resulting in the maintenance ofa constant light level in the tag 30466 inAugust 2004 (top) and tag 43933 in July2004 (bottom). Depth readings are pre-sented in blue and light readings in grey.(b) Detail of a period of diurnal behaviourin tag 43933 (left panel) with depthsduring the day in grey and at night inblack with the proportion of time at depthduring the day (open circles) and night(filled circles) on the right.



Migration and residency patterns

Although only one SBT was observed to migratebetween the Tasman Sea and the spawning grounds,this directly demonstrates connectivity over largespatial scales within the SBT spawning stock. Dailymovement rates and the scale of this migration areconsistent with those observed in other Thunnus spe-cies (Mather et al., 1995; Block et al., 2001, 2005).Furthermore, the movements observed in this studywere consistent with genetic analyses concluding asingle mixed stock (Grewe et al., 1997).

The trigger for movement of SBT away from theirwinter foraging grounds in the Tasman Sea may belinked to seasonal changes in oceanography. Duringthe austral spring, warm waters associated with theEast Australian Current extend down the easternAustralian coastline producing areas of productivitywhich may move suitable foraging areas for SBTsouthward. Seasonal upwelling events also begin in thewaters to the north-west of Tasmania, increasingproductivity in this region and further west into theeastern margins of the GAB throughout spring, sum-mer and early autumn months (Schahinger, 1987;Herzfeld, 1997; Kampf et al., 2004). The area off thesouth-west of Australia is also a region of seasonalproductivity, largely driven by mixing of the southernboundary of the Leeuwin Current waters and coolerSouthern Ocean waters (Pearce and Pattiaratchi,1999). This region supports a number of fisheries ofsmall pelagic species which may serve as an importantforaging area for large SBT. Sub-adult SBT have beenobserved to switch preferred prey species betweenareas of residency (Young et al., 1997) and theirmovements between areas have been hypothesized tobe triggered by a drive to capitalize on concentrationsof small pelagic fishes (Young et al., 1996). Movementof large SBT may not only be driven by reproductiverequirements but also by differential distribution ofprey resources.

It is notable that the position estimates from thisstudy suggest that SBT spend the vast proportion oftheir time on continental shelf areas – 84% of geo-location estimates were within the AFZ. This appearsto be in contrast with juvenile SBT, who regularlymake cross-basin movements into the Indian Ocean(Gunn and Block, 2001), and catch records show thatSBT are caught as far south as 50–55�S (Shingu,1978). A number of factors may contribute to thisperceived high residency in shelf and AFZ waters.First, light-based methods of geolocation are subject toa number of sources of error (Welch and Eveson,1999). Although we were able to compare position

estimates derived using geolocation methods with endpoints derived from Argos locations, determination ofthe accuracy position estimates is somewhat limited.Argos locations are subject to varying degrees of errorthemselves (Hays et al., 2001) and position estimatesdetermined using light and water temperature col-lected at the water surface after a tag has surfacedcannot be considered to be typical of light and tem-perature data collected from the animal. As a result,positions estimates inside the AFZ may actually havebeen outside. However, when considering movementsacross large spatial scales, those data collected fromthe SBT that migrated westward clearly indicateextended residency within the western Tasman Searegion and in shelf areas. Secondly, the attachmentdurations of the PSATs deployed may bias estimatedresidency periods somewhat. With extended attach-ment periods the extent of migrations away from shelfareas would no doubt become clearer.

Considerable variability in dispersal rates andmovement paths resulted in SBT tagged in the samearea being widely dispersed throughout southernAustralian waters during the austral summer. All taggedSBT were within the size range observed in the Indo-nesian spawning ground catches (Farley and Davis,1998) and only one individual was considered not to befully recruited to the spawning stock according tomethods used in previous CCSBT stock assessments.Such wide dispersion, the lack of coherency in move-ments and the duration of residency periods in theSouthern Ocean and Tasman Sea, relative to the timingof spawning ground abundance peaks, hints at the pos-sibility that reproductively mature SBT may not spawnannually. This possibility has been raised in other tunaspecies (Sibert et al., 2006b). However, before definiteconclusions can be made on spawning participationin SBT, information from longer-term deployments isneeded to clearly establish the spatial dynamics ofspawning-sized fish. Furthermore, the two peaksin spawning ground abundance (Farley and Davis, 1998)may indicate variable timing of migrations that maybe due to spatial structuring within the stock. It ispossible that fish observed to be south of Australia whentheir tags detached in late summer could have contin-ued on to spawn. The movement patterns observed inour study are also consistent with other studiessuggesting an older age⁄size at maturity (Farley andDavis, 1998; Schaefer, 2001) than the current CCSBTspawning classification.

The data collected in this study represent a majorstep towards reducing uncertainty about the spatialdynamics of a large over-exploited pelagic predator.At the same time, they raise further questions as to



spawning dynamics and the location of other impor-tant foraging areas for this species in the Australianregion. Despite the limitations of the technology usedto collect these data (limited attachment durations,data transmission⁄reception capacities and geolocationerror), determining such life history aspects would bedifficult without the use of PSATs. Fisheries data areunable to provide detailed movement data and in thecase of adult SBT there have been very few con-ventional tagging data collected from which to infermovements. Deployment of tags in other areasthroughout the range of SBT and across wider tem-poral periods (not only the austral winter) wouldserve to address problems associated with attachmentdurations and also help to define the connectivity ofSBT in different fishery areas such as the Australianand South African regions. Furthermore, continuedcomprehensive data collected across a multi-yearperiod would serve to establish inter-annual vari-ability in SBT habitat preferences and provideimportant inputs into management regimes such asthe spatial management models used in the westernTasman Sea.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the support ofthe skippers and crew of those vessels from whichtagging was carried out, including David Chaffey andThylacine, Tony Forster and Full Force, Phil Pinnisi andSensation, Frank Pirrello and Baroness, Steve Roskellyand Strike Force and Jamie White and Bianca B. PaigeEveson, Mark Green, Klaas Hartmann, KatherineTattersall and Grant West provided operational,logistical and data-processing support. An earlier ver-sion of this manuscript benefited from commentsprovided by Jason Hartog, Tim Davis and JessicaFarley, Molly Lutcavage and an anonymous reviewer.The project was funded by the Australian FisheriesManagement Agency and the Department of Agri-culture, Fisheries and Forestry. All tagging procedureswere conducted in accordance with TasmanianDepartment of Primary Industries, Water and Envi-ronment Ethics Committee approval.

REFERENCES

Block, B.A., Dewar, H., Blackwell, S.B. et al. (2001) Migratorymovements, depth preferences, and thermal biology ofAtlantic bluefin tuna. Science 293:1310–1314.

Block, B.A., Teo, S.L.H., Walli, A. et al. (2005) Electronictagging and population structure of Atlantic bluefin tuna.Nature 434:1121–1127.

Brill, R.W. and Lutcavage, M.E. (2001) Understanding envi-ronmental influences on movements and depth distributionsof tunas and billfishes can significantly improve populationassessments. Am. Fish. Soc. Symp. 25:179–198.

Brill, R., Lutcavage, M., Metzger, G. et al. (2002) Horizontaland vertical movements of juvenile bluefin tuna (Thunnusthynnus), in relation to oceanographic conditions of thewestern North Atlantic, determined with ultrasonic telem-etry. Fish. Bull. 100:155–167.

Bushnell, P.G. and Jones, D.R. (1994) Cardiovascular andrespiratory physiology of tuna – adaptations for support ofexceptionally high metabolic rates. Environ. Biol. Fish40:303–318.

Carey, F.G. and Robison, B.H. (1981) Daily patterns in theactivities of swordfish, Xiphias-Gladius, observed by acoustictelemetry. Fish. Bull. 79:277–292.

Caton, A.E. (1991) Review of aspects of southern bluefin tunabiology, population and fisheries. Inter-Am. Trop. TunaComm. Spec. Rep. 7:181–350.

CCSBT (2004) Report of the Ninth Meeting of the ScientificCommittee, Seogwipo City, Jeju, Republic of Korea, 13–16September 2004.

CCSBT (2006) Report of the Thirteenth Annual Meeting of theCommission, Miyazaki, Japan, 10–13 October 2006.

Clear, N.P., Gunn, J.S. and Rees, A.J. (2000) Direct validationof annual increments in the otoliths of juvenile southernbluefin tuna, Thunnus maccoyii, by means of a large-scalemark-recapture experiment with strontium chloride. Fish.Bull. 98:25–40.

Davis, T.L.O. and Farley, J.H. (2001) Size distribution ofsouthern bluefin tuna (Thunnus maccoyii) by depth on theirspawning ground. Fish. Bull. 99:381–386.

Domeier, M.L., Dewar, H. and Nasby-Lucas, N. (2003) Mor-tality rate of striped marlin (Tetrapturus audax) caught withrecreational tackle. Mar. Freshwater Res. 54:435–445.

Domeier, M.L., Kiefer, D., Nasby-Lucas, N., Wagschal, A. andO’Brien, F. (2005) Tracking Pacific bluefin tuna (Thunnusthynnus orientalis) in the northeastern Pacific with anautomated algorithm that estimates latitude by matchingsea-surface-temperature data from satellites with temperaturedata from tags on fish. Fish. Bull. 103:292–306.

Farley, J.H. and Davis, T.L.O. (1998) Reproductive dynamics ofsouthern bluefin tuna, Thunnus maccoyii. Fish. Bull. 96:223–236.

Farley, J.H. and Davis, T.L.O. (2005) Update on the Lengthand Age Distribution of SBT in the Indonesian Longline Catchon the Spawning Ground. Working Paper CCSBTESC⁄0509⁄16 Presented to the 10th Meeting of theExtended Scientific Committee of the Commission for theConservation of Southern Bluefin Tuna, Taipei, Taiwan,5–8 September 2005.

Golet, W.J., Cooper, A.B., Carnpbell, R. and Lutcavage, M.(2007) Decline in condition of northern bluefin tuna(Thunnus thynnus) in the Gulf of Maine. Fish. Bull.105:390–395.

Grewe, P.M., Elliott, N.G., Innes, B.H. and Ward, R.D. (1997)Genetic population structure of southern bluefin tuna(Thunnus maccoyii). Mar. Biol. 127:555–561.

Gunn, J. and Block, B. (2001) Advances in acoustic, archivaland satellite tagging of tuna. In: Tuna: Physiology, Ecologyand Evolution. B.A. Block & E.D. Stevens (eds) San Diego:Academic Press, pp. 167–224.



Gunn, J. and Young, J. (1999) Environmental determinants ofthe movement of juvenile southern bluefin tuna. In: FishMovement and Migration. D.A. Hancock, D.C. Smith & J.D.Koehn (eds) Bendigo, Victoria: Australian Society for FishBiology, pp. 123–128.

Gunn, J.S., Patterson, T.A. and Pepperell, J.G. (2003) Short-term movement and behaviour of black marlin Makairaindica in the Coral Sea as determined through a pop-up sa-tellite archival tagging experiment. Mar. Freshwater Res.54:515–525.

Gutenkunst, R., Newlands, N., Lutcavage, M. and Edelstein-Keshet, L. (2007) Inferring resource distributions fromAtlantic bluefin tuna movements: an analysis based on netdisplacement and length of track. J. Theor. Biol. 245:243–257.

Hays, G.C., Akesson, S., Godley, B.J., Luschi, P. and Santidrian,P. (2001) The implications of location accuracy for theinterpretation of satellite-tracking data. Anim. Behav.61:1035–1040.

Hearn, W.S., Polacheck, T., Pollock, K.H. and Whitelaw, W.(1999) Estimation of tag reporting rates in age-structuredmulticomponent fisheries where one component hasobservers. Can. J. Fish. Aquat. Sci. 56:1255–1265.

Herzfeld, M. (1997) The annual cycle of sea surface temper-ature in the Great Australian Bight. Prog. Oceanogr. 39:1–27.

Hobday, A.J. and Hartmann, K. (2006) Near real-time spatialmanagement based on habitat predictions for a longlinebycatch species. Fish. Manage. Ecol. 13:365–380.

Horodysky, A.Z. and Graves, J.E. (2005) Application of pop-upsatellite archival tag technology to estimate postrelease sur-vival of white marlin (Tetrapturus albidus) caught on circleand straight-shank (‘‘J’’) hooks in the western NorthAtlantic recreational fishery. Fish. Bull. 103:84–96.

Itoh, T., Kurota, H., Takahashi, N. and Tsuji, S. (2002) Reportof 2001⁄2002 spawning ground surveys – Working PaperCCSBT-SC⁄0209⁄20 Presented to the 7th Meeting of theExtended Scientific Committee of the Commission for theConservation of Southern Bluefin Tuna, Canberra, Austra-lia, 9–11 September 2002.

Kampf, J., Doubell, M., Griffin, D., Matthews, R.L. and Ward,T.M. (2004) Evidence of a large seasonal coastal upwellingsystem along the southern shelf of Australia. Geophys. Res.Lett. 31:L09310. doi: 10.1029/2003GL019221.

Kirby, D.S., Fiksen, O. and Hart, P.J.B. (2000) A dynamicoptimisation model for the behaviour of tunas at oceanfronts. Fish. Oceanogr. 9:328–342.

Kitagawa, T., Nakata, H., Kimura, S. and Tsuji, S. (2001)Thermoconservation mechanisms inferred from peritonealcavity temperature in free-swimming Pacific bluefin tunaThunnus thynnus orientalis. Mar. Ecol. Prog. Ser. 220:253–263.

Kitagawa, T., Sartimbul, A., Nakata, H., Kimura, S., Yamada,H. and Nitta, A. (2006) The effect of water temperature onhabitat use of young Pacific bluefin tuna Thunnus orientalis inthe East China Sea. Fish. Sci. 72:1166–1176.

Lam, C.H., Nielsen, A. and Sibert, J.R. (2008) Improving lightand temperature based geolocation by unscented Kalmanfiltering. Fish. Res. 91:15–25.

Laslett, G.M., Eveson, J.P. and Polacheck, T. (2002) A flex-ible maximum likelihood approach for fitting growthcurves to tag-recapture data. Can. J. Fish. Aquat. Sci.59:976–986.

Laurs, R.M. and Lynn, R.J. (1977) Seasonal migration ofNorth Pacific albacore, Thunnus-Alalunga, into NorthAmerican coastal waters – distribution, relative abun-dance, and association with transition zone waters. Fish.Bull. 75:795–822.

Lutcavage, M.E., Brill, R.W., Skomal, G.B., Chase, B.C. andHowey, P.W. (1999) Results of pop-up satellite tagging ofspawning size class fish in the Gulf of Maine: do NorthAtlantic bluefin tuna spawn in the mid-Atlantic? Can. J.Fish. Aquat. Sci. 56:173–177.

Lutcavage, M.E., Brill, R.W., Skomal, G.B., Chase, B.C.,Goldstein, J.L. and Tutein, J. (2000) Tracking adult NorthAtlantic bluefin tuna (Thunnus thynnus) in the north-western Atlantic using ultrasonic telemetry. Mar. Biol.137:347–358.

Marcinek, D.J., Blackwell, S.B., Dewar, H. et al. (2001) Depthand muscle temperature of Pacific bluefin tuna examinedwith acoustic and pop-up satellite archival tags. Mar. Biol.138:869–885.

Mather, F.J., III, Mason, J.M., Jr and Jones, C.A. (1995) His-torical document: life history and fisheries of Atlantic bluefintuna. NOAA Tech. Memo. NMFS-SEFSC-370. 1–165.

Musyl, M.K., Brill, R.W., Boggs, C.H., Curran, D.S., Kazama,T.K. and Seki, M.P. (2003) Vertical movements of bigeyetuna (Thunnus obesus) associated with islands, buoys, andseamounts near the main Hawaiian Islands from archivaltagging data. Fish. Oceanogr. 12:152–169.

Nakano, H., Matsunaga, H., Okamoto, H. and Okazaki, M.(2003) Acoustic tracking of bigeye thresher shark Alopiassuperciliosus in the eastern Pacific Ocean. Mar. Ecol. Prog.Ser. 265:255–261.

Neill, W.H., Chang, R.K.C. and Dizon, A.E. (1976) Magnitudeand ecological implications of thermal inertia in skipjacktuna, Katsuwonus pelamis (Linnaeus). Environ. Biol. Fish1:61–80.

Olson, D.B. (2001) Biophysical dynamics of western transitionzones: a preliminary synthesis. Fish. Oceanogr. 10:133–150.

Pearce, A. and Pattiaratchi, C. (1999) The Capes Current: asummer countercurrent flowing past Cape Leeuwin andCape Naturaliste, Western Australia. Cont. Shelf Res.19:401–420.

PO.DAAC (2001) AVHRR Orbital 9km MCSST Level 2(NAVOCEANO). Physical Oceanography Distributed ActiveArchive Center (PO.DAAC) at the NASA. Pasadena, CA: JetPropulsion Laboratory. http://podaac.jpl.nasa.gov.

Polovina, J.J., Howell, E., Kobayashi, D.R. and Seki, M.P.(2001) The transition zone chlorophyll front, a dynamicglobal feature defining migration and forage habitat formarine resources. Prog. Oceanogr. 49:469–483.

Preece, A., Cooper, S. and Hartog, J. (2004) Data Post-processingfor Input to the 2004 Stock Assessments and Comparisons of2001 and 2004 Assessment Datasets Working paper CCSBT-ESC⁄0409⁄27, Seogwipo City, Jeju, Republic of Korea, 13–16September 2004.

Royer, F., Fromentin, J.M. and Gaspar, P. (2004) Associationbetween bluefin tuna schools and oceanic features in thewestern Mediterranean. Mar. Ecol. Prog. Ser. 269:249–263.

Royer, F., Fromentin, J.M. and Gaspar, P. (2005) A state-spacemodel to derive bluefin tuna movement and habitat fromarchival tags. Oikos 109:473–484.

Schaefer, K.M. (2001) Reproductive biology of tunas. In: TunaPhysiology, Ecology, and Evolution. B.A. Block & E.D. Ste-vens (eds) New York: Academic Press, pp. 225–270.



Shaffer, S.A., Tremblay, Y., Awkerman, J.A., Henry, R.W., Teo,S.L.H., Anderson, D.J., Croll, D.A., Block, B.A. and Costa,D.P. (2005) Comparison of light-and SST-based geolocationwith satellite telemetry in free-ranging albatrosses. MarineBiology, 147:833–843.

Schahinger, R.B. (1987) Structure of Coastal Upwelling EventsObserved off the Southeast Coast of South-Australia duringFebruary 1983–April 1984. Aust. J. Mar. Freshwater Res.38:439–459.

Schick, R.S., Goldstein, J. and Lutcavage, M.E. (2004) Bluefintuna (Thunnus thynnus) distribution in relation to sea surfacetemperature fronts in the Gulf of Maine (1994–96). Fish.Oceanogr. 13:225–238.

Shingu, C. (1978). Ecology and stock of southern bluefin tuna.Fish Study Jap. Assoc. Fish. Resources Protection, 31:81 p. (InJapanese; English translation in Rep. CSIRO Div.Fish.Oceanogr., 131:79 p., 1981.).

Sibert, J., Hampton, J., Kleiber, P. and Maunder, M. (2006a)Biomass, size, and trophic status of top predators in thePacific Ocean. Science 314:1773–1776.

Sibert, J.R., Lutcavage, M.E., Nielsen, A., Brill, R.W. andWilson, S.G. (2006b) Interannual variation in large-scalemovement of Atlantic bluefin tuna (Thunnus thynnus)determined from pop-up satellite archival tags. Can. J. Fish.Aquat. Sci. 63:2154–2166.

Stokesbury, M.J.W., Teo, S.L.H., Seitz, A., O’Dor, R.K. andBlock, B.A. (2004) Movement of Atlantic bluefin tuna(Thunnus thynnus) as determined by satellite tagging exper-iments initiated off New England. Can. J. Fish. Aquat. Sci.61:1976–1987.

Sund, P.N., Blackburn, M. and Williams, F. (1981) Tunas andtheir environment in the Pacific Ocean: a review. Oceanogr.Mar. Biol. Annu. Rev. 19:443–512.

Teo, S.L.H., Boustany, A., Blackwell, S., Walli, A., Weng, K.C.and Block, B.A. (2004) Validation of geolocation estimatesbased on light level and sea surface temperature from elec-tronic tags. Mar. Ecol. Prog. Ser. 283:81–98.

Teo, S.L.H., Boustany, A., Dewar, H. et al. (2007) Annualmigrations, diving behavior, and thermal biology of Atlanticbluefin tuna, Thunnus thynnus, on their Gulf of Mexicobreeding grounds. Mar. Biol. 151:1–18.

Toscas, P.J., Venables, W.N., Thomas, M. and Polacheck, T.(2001) Modelling Catch and Effort in the Southern Bluefin TunaFishery. Working Paper CCSBT-CPUE⁄0203⁄BGD⁄02 Pre-sented to the 6th Meeting of the Extended Scientific Com-mittee of the Commission for the Conservation of SouthernBluefin Tuna, Tokyo, Japan, 28–31 August 2001.

Venables, W.N. and Ripley, B.D. (1997) Modern Applied Sta-tistics with S-Plus. Berlin: Springer.

Warrant, E. (2000) The eyes of deep-sea fishes and the changingnature of visual scenes with depth. Philos. Trans. R. Soc.Lond., B, Biol. Sci. 355:1155–1159.

Welch, D.W. and Eveson, J.P. (1999) An assessment of light-based geoposition estimates from archival tags. Can. J. Fish.Aquat. Sci. 56:1317–1327.

Weng, K.C. and Block, B.A. (2004) Diel vertical migrationof the bigeye thresher shark (Alopias superciliosus), a speciespossessing orbital retia mirabilia. Fish. Bull. 102:221–229.

Wilson, S.G., Lutcavage, M.E., Brill, R.W., Genovese, M.P.,Cooper, A.B. and Everly, A.W. (2005) Movements ofbluefin tuna (Thunnus thynnus) in the northwestern AtlanticOcean recorded by pop-up satellite archival tags. Mar. Biol.146:409–423.

Young, J.W., Bradford, R.W., Lamb, T.D. and Lyne, V.D. (1996)Biomass of zooplankton and micronekton in the southernbluefin tuna fishing grounds off eastern Tasmania, Australia.Mar. Ecol. Prog. Ser. 138:1–14.

Young, J.W., Lamb, T.D., Le, D., Bradford, R.W. and Whitelaw,A.W. (1997) Feeding ecology and interannual variations indiet of southern bluefin tuna, Thunnus maccoyii, in relation tocoastal and oceanic waters off eastern Tasmania, Australia.Environ. Biol. Fish 50:275–291.



Date post:	17-Aug-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Movement and behaviour of large southern bluefin tuna ... · tional double-barbed nylon dart tip...

Documents